84 AI-personas reveal how Belgians truly connect with Colruyt Group’s core values
Challenge
They reached out to us to evaluate a previous internal project involving synthetic testing. Although Colruyt had already tried synthetic testing internally, they hadn’t achieved the desired results and had decided to discontinue the process. When they contacted us, their question was simple: "Why would this work with you?"
To address their concerns, we proposed a comparative test. We took a study Colruyt had previously conducted with real respondents (with the results unknown to us) and replicated it using synthetic methods, combining both quantitative and qualitative methodologies.
The outcomes of our synthetic testing were closely aligned with Colruyt’s original data, and in some areas, we were able to explore deeper insights. Impressed by these results, the Colruyt Group decided to move forward with a follow-up project: evaluating the credibility of their three core pillars with different target audiences. The second project wasn’t about solving a problem in the traditional sense – it was about seizing an opportunity to push innovation further. By using synthetic testing, Colruyt aimed to ensure their strategies aligned perfectly with what their consumers truly valued, showcasing their commitment to continuous improvement.
The Approach
For this project, we worked with 84 synthetic personas representing the diverse and engaged Belgian population aged 18-75. The reason for this is the equally broad target audience that Colruyt Group reaches through their brand portfolio. These AI-modeled personas are designed to mimic the exact same human behavior of the target audience, allowing us to gather accurate insights from individuals who regularly visit Colruyt Group stores and directly influence purchasing decisions.
A key advantage of using synthetic personas is that it removes the risk of human bias, which could arise from prior experiences with survey subjects, whether positive or negative. Additionally, using synthetic personas allowed us to conduct multiple iterations of follow-up questions, something not possible with traditional qualitative interviews where the number of respondents is limited. This enabled us to refine our understanding further and dive deeper into specific areas of interest, providing more comprehensive insights.
We divided the synthetic personas into two groups:
- One group received only the core pillars with no additional context.
- The second group received detailed information about the initiatives Colruyt undertakes to realize these pillars.
For each core pillar, respondents were asked four questions, totaling 12 across all pillars. These questions were designed to assess their associations with each topic and their perception of how Colruyt Group aligns with them.
The Result
While we can’t share all the details of the study results, both studies delivered powerful insights that can serve as a strong foundation for Colruyt Group’s future business strategies.
First, we validated the added value of synthetic testing for Colruyt Group. The most significant benefits for them were that synthetic testing is less time-consuming and more budget-friendly compared to traditional surveys. Moreover, synthetic testing provided deeper, more objective insights by offering a broader and more accurate understanding of customer behaviors and preferences.
The results from the second study provided large-scale insights that can serve as a foundation for Colruyt Group’s future development. These insights can also be cross-validated with real people, should they wish to do so, helping them refine their strategies. Furthermore, we didn’t just deliver the analysis and results—we worked with Colruyt to iterate on what these insights mean for their strategy and how they can be effectively integrated into their communication efforts, ensuring the right messages and core pillars resonate with their target audience.
